Using Fixed Effects Stata and Random Effects

[Roberto Villa]

Dear all

I am getting an error when I use the follow fixed effects codes.

• I have a balanced panel data. (Individual= state and across 4 time periods)
• I have two fixed effects : Individual State Fixed Effect= total panel observations/ year and time fixed effect (year)

when I use

lets say
k=state
t=years

xtreg y x1 x2 x3 x4 , fe i(k t)

i get the following error : i(): too many variables specified

Also,

areg y x1 x2 x3 x4 , absorb (k t )

i get the following error : absorb(): too many variables specified

Questions

Q1 Basically my question is ; how to use two way fixed effects and random effects??? with
Fixed Effects
xtreg , fe
areg ,absorb

Random Effects
xtmixed || :,


Q2 should I just include i.year dummy after (xtreg ,areg xmixed) and before ( , fe , absorb , || :.)

Thank you

[Carlo Lazzaro]

Roberto:
what f you go:

xtset panelid timevar
xtreg y x1 x2 x3 x4 i.timevar, fe